基于多模PPP-AR技术的大气水汽探测性能分析

精密单点定位（precise point positioning，PPP）反演大气可降水量（precipitable water vapor，PWV）具有精度高、实时性强等优点，能够在灾害监测、降雨预报及探测降水信息等方面发挥重要作用。为评估整周模糊度固定模式下PPP-AR（PPP ambiguity resolution）反演PWV的性能，选取全球范围16个MGEX站2022年4个时段的观测数据，采用最终精密星历解算，设置不同星座组合（GPS,BDS-3,GPS+BDS-3,GPS+GLO+GAL+BDS-3）获取对流层延迟（zenith total delay，ZTD）估值，并转换为PWV。从PPP-ZTD与IGS-ZTD的相关性、PPP-ZTD收敛时间、ZTD估值精度和PPP-PWV估值精度4个方面评价多模PPP-AR探测水汽的性能。结果表明，与单（G、C）、双系统（GC）固定解相比，多系统（GREC）固定解获取ZTD估值更加精确，相关系数更高。相较于单、双系统，多系统具有更快的收敛速度，收敛时间分别缩短了27%,25%和20%，多系统固定解与浮点解相比收敛时间缩短11%。此外，对GNSS PPP反演的PWV与探空站PWV（RS-PWV）进行对比，结果表明，WUH2站与HOB2站单、双、多系统固定解、多系统浮点解（float-GREC）的平均均方根误差分别为6.40 mm,6.48 mm,6.19 mm,6.17 mm,6.19 mm和5.82 mm,5.77 mm,5.72 mm,5.62 mm,5.70 mm。多模下得到的PWV估值精度最高，可为高精度的水汽反演提供支持。

Performance analysis of atmospheric water vapor detection based on multi-constellation PPP-AR technology

Precise point positioning (PPP) to retrieve precipitable water vapor (PWV) has the advantages of high accuracy, strong real-time performance, and can play an important role in terms of disaster monitoring, rainfall forecasting and detecting rainfall information. In order to evaluate the performance of PPP ambiguity resolution (PPP-AR) inversion of PWV, observation data from 16 MGEX stations around the world for four periods in 2022 are selected, and the final precise ephemeris is used for calculation, and different constellation combinations (GPS, BDS-3, GPS+BDS-3, GPS+GLO+GAL+BDS-3) are set to obtain zenith total delay (ZTD) valuation and the inversion of PWV. The water vapor detection performance of multi-constellation PPP-AR is evaluated from four aspects: the correlation between PPP-ZTD and IGS-ZTD, the convergence time of PPP-ZTD, the estimation accuracy of ZTD and the estimation accuracy of PPP-PWV. The results show that compared with single(G, C) and dual system (GC) fixed solutions, the multi-constellation (GREC) fixed solution obtains ZTD estimation more accurately and has a higher correlation coefficient. Compared with single and dual systems, the multi-constellation has faster convergence speed, and the convergence time is shortened by 27%, 25% and 20% respectively. The multi-constellation fixed solution has a convergence time shortened by 11% compared with the floating-point solution. In addition, the comparison of PWV and Radiosondes PWV (RS-PWV) retrieved by GNSS PPP shows that the mean root mean square errors of the single system (G, C), dual system (GC), multi-constellation fixed solution (GREC) and multi-constellation floating-point solution (float-GREC) of the WUH2 station are 6.40 mm, 6.48 mm, 6.19 mm, 6.17 mm, 6.19 mm, respectively; and those of the HOB2 station are 5.82 mm, 5.77 mm, 5.72 mm, 5.62 mm and 5.70 mm, respectively. The PWV estimation accuracy obtained in multi-constellation is the highest, which can support high-precision water vapor inversion.